Implementing the data stack [Checklist]

This document was prepared by Justin Butlion of  SaaS Decoded . It provides a step-by-step checklist to implement the modern data stack shown in the diagram below.





Add a caption...

Sign up to Fivetran

Invite CTO / COO to Fivetran

Invite analysts / ops people to Fivetran

Connect application database to Fivetran

Set the sync to once a day in the early hours of the morning (3AM local time for example)

Connect the subscription management solution to Fivetran

Add a credit card to Fivetran

Set up usage alerts

Sign up to Google Cloud Platform

Create a project for the data stack

Set up billing account for the project

Add credit card to the billing account

Add CTO / COO to Google Cloud Platform

Add analysts / ops people to Google Cloud Platform

Turn on BigQuery API

Add budget alerts to the billing account

Create a repository in Github dedicated to dbt. If you don't use Github, create a repository in whichever code repository solution you use. dbt works with Github, Gitlab and Azure DevOps. There is also a self-hosted option which I don't recommend.

Invite whoever will be building models in dbt to Github and add them as admins on the new repository

Sign up to dbt

Connect previously created repository to dbt

Invite analyst / data engineer / developer to dbt (will need to purchase at least 1 developer license)

Connect BigQuery to dbt

Make sure that you can read data in BigQuery

Sign up to Segment

Create a workspace for dev

Invite CTO to both the production and dev workspaces

Add your backend as a source in both workspaces

Share the write keys with the CTO

Create Google sheet for your tracking plan

Fill in the tracking plan

Sync with the CTO to build a plan to develop the events

Connect BigQuery as a destination

QA the events once they go live in the dev environment

Push the events to production and confirm everything is in order

Sign up to Mixpanel

Connect Mixpanel to Segment